RE: [GNUnet-developers] Reproduceable failure

[Jan Marco Alkema]

Hello Igor, Tracy,

>My gnunetd dies quite regularly. It won't run for more than an hour or two.

> In addition my gnunetd dies even more regularly by "Killed"

My gnunetd process on a pentium 4 1700 Mhz 256 Mb RIMM mem (Redhat 8.0):

# ./gnunet-stats
Uptime (seconds) :  570983 => 159 hours => 6.60 days uptime!

>because the task ate too much memory for the kernel's liking.

Someone told me to use the source code of the 'top' shell command for mem
monitoring of the gnunetd process.

>Mar 31 04:52:42 __BREAK__

I think it is better to patch gdb in such a way that automatically after a
segmentation error occurs a dump is made and the restart of gnunetd will be
done. After restarting of gnunetd the "info threads dump" can be send to the
maintainer (for example by the chat client of gnunet).

Suggestion: If in gnunetd a __BREAK__ exception has occurred, gnunetd must
"signal" to the patched gdb process to do an "info threads dump" at the
place of where the __BREAK__ is made in the source code of gnunet,

Greetings Jan Marco